This directory contains a codebook for all the variables in Ying's "Military Power and Ideological Appeals of Religious Extremists."


*** Magazine Analysis (main dataset: 1_magazine.csv) ***

id: Unique identifier for each page.
group: Name of the violent group.
magazine_name: Name of the magazine.
magazine: Name of the magazine. Unlike "magazine_name," this variable provides extensions to distinguish between different magazines that share the same English name.
issue: Magazine issue; each magazine generally includes multiple issues.
language: Language in which the magazine is written.
time: Release time of the magazine issue.
posttime2mon: Two months after the magazine issue's release date.
pretime3mon: Three months after the magazine issue's release date.
pretime2mon: Two months before the magazine issue's release date.
type: Type of propaganda, either "Magazine" or "Newsletter."
total_issue: Total number of issues for a specific magazine.
women: Indicates if the magazine is targeted at women.
rawtime: Unprocessed time variable as collected.
year: Year the magazine issue was released.
religscore: Religiosity score of the magazine page based on the text (number of religious words minus the number of secular words).
frontpage: Indicates whether the page is the front page.
westlang: Indicates whether the magazine is in a Western language.
posscore: Positivity score of the issue, measured by the LIWC dictionary (number of positive words minus the number of negative words).
competitors: Number of competing violent groups in the same country.
page: Specific page number within the current magazine issue.
power: Language about "power," measured by the LIWC dictionary.
sad: Language about "sad," measured by the LIWC dictionary.
death: Language about "death," measured by the LIWC dictionary.
totalevent3mon: Total number of violent events conducted by the group over the three months before the current magazine page was published.
statebased3mon: Total number of battles against government forces conducted by the group over the three months before the current magazine page was published.
totaldeaths3mon: Total number of deaths caused by the group over the three months before the current magazine page was published.
totalevent2mon: Total number of violent events conducted by the group over the two months before the current magazine page was published.
statebased2mon: Total number of battles against government forces conducted by the group over the two months before the current magazine page was published.
totaldeaths2mon: Total number of deaths caused by the group over the two months before the current magazine page was published.
civildeaths2mon: Total number of civilian deaths over the two months before the current magazine page was published.
totalevent2monpost: Total number of violent events conducted by the group over the two months after the current magazine page was published.
statebased2monpost: Total number of battles against government forces conducted by the group over the two months after the current magazine page was published.
totaldeaths2monpost: Total number of deaths caused by the group over the two months after the current magazine page was published.
pagenum: Total number of pages in the current magazine issue.



*** Twitter Analysis (main dataset 1: 1_twitter_tweets.csv) ***

created_date: Date the tweet was created.
tweet_id: Unique identifier for each tweet.
created_time: Exact time the tweet was created.
user_id: Unique identifier for the user who posted the tweet.
screen_name: User's Twitter handle or screen name.
user_created_time: Exact time the user's Twitter account was created.
user_created_date: Date the user's Twitter account was created.
followers_count: Number of followers the user has at the time of the tweet.
favourites_count: Number of tweets the user has liked.
friends_count: Number of accounts the user is following.
statuses_count: Total number of tweets the user has posted.
location: User-provided location information.
time_zone: User-provided time zone information.
hashtags: List of hashtags used in the tweet.
media: Indicates whether links to media (images, videos, etc.) are included in the tweet.
urls: Indicates whether any URLs are included in the tweet.
mentions_ids: IDs of users mentioned in the tweet.
mentions_screen_names: Screen names of users mentioned in the tweet.
daystoholi: Number of days remaining until the nearest Islamic holiday.
mentioncore: Number of core ISIS members mentioned in the tweet.
mentionmember: Number of ISIS members mentioned in the tweet.
mentiontotal: Total number of accounts mentioned in the tweet.
all_retweet: Total number of retweets.
members_retweet: Retweets made specifically by ISIS members.
core_retweet: Retweets made specifically by ISIS core members (defined by popularity).
geo_retweet: Total number of retweets from which geographic location can be inferred.
geo_core_retweet1: Retweets made specifically by ISIS core members, defined geographically (users in certain time zones such as Iraq, Syria, Qatar, Saudi Arabia, and Turkey are classified as core members; those in other regions are classified as peripheral).
daily_active_user: Daily count of active users tweeting.
daily_avg_followers: Average number of followers per day.
daily_active_core: Daily count of active core members tweeting.
daily_herf: Herfindahl index, measuring the concentration of tweets by geography.
territory: Area of territory controlled by ISIS.
territory_lag_10: Area of territory controlled by ISIS with a ten-day lag.
totalevent1: Total number of violent events conducted by ISIS the day before the tweet.
statebased1: Total number of battles against government forces conducted by ISIS the day before the tweet.
totaldeaths1: Total number of deaths caused by ISIS the day before the tweet.
religscore: Religiosity score of the tweet's content.
accountlength: Time since the user's account was created.
mentionout: Number of accounts outside the ISIS network mentioned in the tweet.
Imentionout: Whether the tweet mentions any people outside of the ISIS network.
Ihashtags: Whether the tweet contains hashtags.
daily_core_prop: Daily proportion of core members.


*** Twitter Analysis (main dataset 2: 1_twitter_panel.csv) ***

user_id: Unique identifier for the user who posted the tweet.
created_date: Date the tweets were created.
meanscore: Average religiosity score of tweets by the user on the given date.
territory: Area of territory controlled by ISIS on the given date.
meanmention: Average "Imentionout" of the day.
totalevent1: Total number of violent events conducted by ISIS on the previous day.
daystoholi: Number of days remaining until the nearest Islamic holiday.
meanfollowers: Average number of followers the user has on the given date.
meanfavourites: Average number of tweets liked per tweet by the user on the given date.
meanfriends: Average number of accounts the user is following on the given date.
meanstatuses: Average total number of tweets posted by the user on the given date.
meanaccountlength: Average account age  by the user on the given date.
meanmedia: Average presence of media links tweeted by the user on the given date.
meanurls: Average presence of URLs tweeted by the user on the given date.
meanhashtags: Average use of hashtags by the user on the given date.
mean_core_perip: Average measure of retweets by core members minus that by peripheral members.
mean_memb_out: Average number of retweets by ISIS members minus that by non-members.
mean_geo_core1_noncore1: Average number of retweets by geographically core members minus that by non-core members.
mean_members_retweet: Average number of retweets by ISIS members.
mean_out_retweet: Average number of retweets by non-members.
mean_core_retweet: Average number of retweets by core members.
mean_perip_retweet: Average number of retweets by peripheral members.
mean_geocore1_retweet: Average number of retweets by geographically core members.
mean_geoperip1_retweet: Average number of retweets by geographically non-core members.
daily_active_user: Daily count of active users tweeting.
daily_core_prop: Daily proportion of core members.
daily_herf: Herfindahl index, measuring the concentration of tweets by geography.
ntweets: Total number of tweets posted by the user on the given date.



